🎉New preprint! @LauraKremer20 and I looked at four different gene knockouts in mt.5024C>T mice to investigate how #autophagy affects #mtDNA heteroplasmy inheritance. biorxiv.org/content/10.110…
@MRC_MBU @cam_mito @mitopinionated 🧵: 1/
@MRC_MBU @cam_mito @mitopinionated 🧵: 1/
To what extent is germline purifying selection driven by autophagy? We looked at the mother-to-pup heteroplasmy shift in high-heteroplasmy mice carrying Parkin, Bcl2l13, Ulk1, and Ulk2 knockouts, as well as two separate control populations. 2/ 
While previously Filograna et al. (2019) @RobertaFilogra2 show TFAM overexpression can lead to higher maximal tolerated heteroplasmy, this was not the case for our autophagy knockouts. We saw no knockout mice with H>80%, despite having 120+ pups in each group. 3/
However, at a first glance we do see a change in heteroplasmy shift (Burgstaller et al., 2014). There appears to be purifying selection in the controls (as expected) and Parkin knockout group, but perhaps less so in the knockout of Bcl2l13, Ulk1, and Ulk2. 4/
But is this effect *really there* for Bcl2l13, Ulk1, and Ulk2? Despite having 120+ pups, most genotype groups have <7 mothers. Measurement error or unrelated sources of variability in the mothers might be leading to spurious "significance"! 5/
To account for this, we resampled the control data, considering different combinations of control mothers to test against. The Bcl2l13 knockout effect is observed regardless of how we pick our controls. However, Ulk1 and Ulk2 effects are only seen sometimes. 6/
Afterwards we asked: is the Bcl2l13 effect so clear just because there’s more data in this group (236 pups vs <160 in other groups)? Not the case! We can reliably detect the Bcl2l13 effect even after subsampling down to 120 pups. 7/
🚨Overall, we show robust evidence that Bcl2l13, a homolog of the yeast #mitophagy receptor Atg32, plays a role in purifying #mtDNA selection. 8/
We hypothesise that Ulk1 and Ulk2 might play a role in selection as well. One reason we might struggle seeing a robust effect is that the two genes are homologous and partly functionally redundant, while a double knockout is lethal. 9/
Finally, we've made concerted effort to ensure our work is transparent and fully #reproducible, which I’m personally very proud of! 🥰 #rstats code: github.com/lbozhilova/bcl… #OpenScience 🧵10/10
• • •
Missing some Tweet in this thread? You can try to
force a refresh
